Data security management method and device, computer program product and electronic equipment

Abstract

The invention discloses a data security management method and device, a computer program product and electronic equipment. The method comprises the steps of obtaining a first data set, related to a target field, of a target object in a target time period and a personalized privacy protection requirement of the target object; analyzing each piece of first data in the first data set and the description information thereof by using a target data sensitivity analysis model, and determining the sensitivity level of each piece of first data; a target privacy protection strategy of each first data is decided according to the personalized privacy protection requirement and the sensitivity level of each first data; and for each piece of first data, encrypting the first data according to the target privacy protection strategy, and storing the encrypted first data in the block chain network. The technical problem that when a related data management system carries out safety management on data, the data privacy protection requirement of a user is not considered, so that the trust degree and the acceptability of the user to the data management system are low is solved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and more specifically, to a data security management method and device, a computer program product, and an electronic device. Background Art

[0002] In the current digital environment, data security and privacy protection have become crucial issues. Blockchain technology is widely regarded as an ideal solution for strengthening data security and privacy protection due to its inherent decentralization, unalterable ledger records and strong encryption measures.

[0003] The existing blockchain technology solutions in the field of data security and privacy protection mainly include the following aspects: encryption technology, smart contracts, consensus mechanism, and privacy enhancement technology. However, when relevant data management systems use blockchain technology to manage data privacy, they usually set corresponding security protection measures based on the data itself, without considering the user's privacy protection needs for the data, which greatly reduces the user's trust and acceptance of the data management system.

[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0005] The embodiments of the present application provide a data security management method and device, a computer program product, and an electronic device to at least solve the technical problem that when the relevant data management system performs security management of data, it does not consider the user's data privacy protection needs, resulting in low user trust and acceptance of the data management system.

[0006] According to one aspect of an embodiment of the present application, a data security management method is provided, including: obtaining a first data set related to a target field of a target object within a target time period, and obtaining personalized privacy protection requirements of the target object, wherein the first data set includes first data of the target object at multiple moments within the target time period and description information of each first data; using a target data sensitivity analysis model to analyze each first data and its description information in the first data set, and determine the sensitivity level of each first data in the first data set, wherein the data sensitivity analysis model is obtained by fine-tuning a pre-trained initial data sensitivity analysis model using a corpus of the target field, and the initial data sensitivity analysis model is obtained by training a neural network model with a general corpus; determining a target privacy protection strategy for each first data according to the personalized privacy protection requirements of the target object and the sensitivity level of each first data in the first data set; for each first data, encrypting the first data according to the target privacy protection strategy, and storing the encrypted first data on a blockchain network.

[0007] Optionally, obtaining the personalized privacy protection requirements of the target object includes: obtaining the personalized privacy protection requirements of the target object includes: in response to multiple privacy protection parameter setting instructions initiated by the target object on the privacy protection setting interface, using the privacy protection parameters corresponding to the multiple privacy protection parameter setting instructions as the personalized privacy protection requirements of the target object, wherein the personalized privacy protection requirements include at least one of the following: privacy protection level, data access control, data retention period, and data sharing scope.

[0008] Optionally, before using the target data-sensitive analysis model to analyze each first data and its description information in the first data set, the method also includes: determining the data type of each first data in the first data set, wherein the data type includes at least one of the following: text type, voice type, image type; when the type of the first data is text type data, using a first preprocessing operation to preprocess the first data, wherein the first preprocessing operation includes at least one of the following: text cleaning, case conversion, word segmentation, stop word removal, stem extraction, word form restoration; when the type of the first data is voice type, using a second preprocessing operation to preprocess the first data, wherein the second preprocessing operation includes at least one of the following: frame processing, noise elimination, voice recognition; when the type of the first data is image type, using a third preprocessing operation to preprocess the first data, wherein the third preprocessing operation includes at least one of the following: resizing, data enhancement, normalization.

[0009] Optionally, when the type of the first data is text type data or speech type, the model architecture of the initial data sensitive analysis model includes at least one of the following: a convolutional neural network, a long short-term memory network, a recurrent neural network, a Transformer model based on an attention mechanism; when the type of the first data is image type data, the model architecture of the initial data sensitive analysis model includes: a deep residual network or a densely connected convolutional network.

[0010] Optionally, a target privacy protection strategy for each first data is decided based on the personalized privacy protection needs of the target object and the sensitivity level of each first data in the first data set, including: for each first data in the first data set, using a pre-trained target decision model to analyze the sensitivity level of the first data and the personalized privacy protection needs of the target object to obtain the target privacy protection strategy for the first data, wherein the target privacy protection strategy includes at least one privacy protection technology, and the privacy protection technology includes at least one of the following: zero-knowledge proof technology, homomorphic encryption technology, secure multi-party computing technology, differential privacy technology, and proxy re-encryption technology.

[0011] Optionally, the training process of the target decision model includes: obtaining a training sample set and a sample label set, wherein the training sample set includes multiple training samples consisting of sensitivity level labels corresponding to historical data related to the target field by other users at historical moments and personalized privacy protection requirements of other objects that uploaded historical data, and the sample label set includes optimal privacy protection strategy labels corresponding to each training sample; constructing an initial decision model, wherein the initial decision model includes at least one of the following: a decision tree model, a logistic regression model, and a support vector machine model; and iteratively training the initial decision model using the training sample set and the sample label set to obtain the target decision model.

[0012] Optionally, encrypting the first data according to the target privacy protection strategy and storing the encrypted first data on the blockchain network includes: encrypting the first data according to the target privacy protection strategy, and packaging the encrypted first data into transaction data, wherein the transaction data at least includes: the first data, the transaction timestamp, and the public key of the target object; signing the transaction data using the public key of the target object, and broadcasting the signed transaction data to each node in the blockchain network; for each node on the blockchain network, verifying the signed transaction data using a consensus mechanism, and adding the verified transaction data to the block to be packaged, wherein the block to be packaged also includes other transaction data and a hash value for linking to the previous block; packaging the blocks to be packaged into a new block, and after the new block is confirmed by other nodes on the blockchain network through a consensus mechanism, adding the new block to the blockchain network to update the ledgers of all nodes in the blockchain network.

[0013] According to another aspect of an embodiment of the present application, a data security management device is also provided, including: an acquisition module, used to acquire a first data set related to a target field of a target object within a target time period, and acquire personalized privacy protection requirements of the target object, wherein the first data set includes first data of the target object at multiple moments within the target time period and description information of each first data; an analysis module, used to analyze each first data and its description information in the first data set using a target data sensitivity analysis model, and determine the sensitivity level of each first data in the first data set, wherein the data sensitivity analysis model is obtained by fine-tuning a pre-trained initial data sensitivity analysis model using a corpus of the target field, and the initial data sensitivity analysis model is obtained by training a neural network model with a general corpus; a decision module, used to decide a target privacy protection strategy for each first data according to the personalized privacy protection requirements of the target object and the sensitivity level of each first data in the first data set; a security protection module, used to encrypt each first data according to the target privacy protection strategy, and store the encrypted first data on a blockchain network.

[0014] According to another aspect of an embodiment of the present application, a computer program product is further provided, the computer program product comprising: a computer program, wherein the computer program implements the above-mentioned data security management method when executed by a processor.

[0015] According to another aspect of an embodiment of the present application, an electronic device is further provided, comprising: a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to execute the above-mentioned data security management method through the computer program.

[0016] In the embodiment of the present application, the management system uses integrated learning combined with deep learning models and machine learning algorithms to conduct in-depth analysis and sensitivity identification of data, thereby achieving flexibility and customization of privacy protection measures. At the same time, users can dynamically adjust privacy protection measures according to their personal privacy preferences. This flexibility allows users to choose appropriate privacy protection technologies when dealing with different types of transactions, thereby effectively meeting individual needs. This highly customized solution enhances the user's control over privacy protection and improves user satisfaction. This solves the technical problem that when the relevant data management system performs security management of data, it does not consider the user's data privacy protection needs, resulting in low user trust and acceptance of the data management system. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0018] Figure 1 is a flowchart of an optional data security management method according to an embodiment of the present application;

[0019] Figure 2 is a schematic structural diagram of an optional data security management device according to an embodiment of the present application;

[0020] Figure 3 It is a schematic diagram of the structure of an optional electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0021] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0022] It should be noted that the terms "first", "second", etc. in the specification, claims and drawings of the present application are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0023] In order to better understand the embodiments of the present application, some nouns or terms that appear in the description of the embodiments of the present application are first translated and explained as follows:

[0024] Blockchain: It is a decentralized distributed ledger that is tamper-proof, secure and reliable. It combines distributed storage, peer-to-peer storage, consensus mechanism, cryptography and other technologies to record transactions and information through a growing data block chain to ensure data security and transparency.

[0025] Smart Contract: A computer protocol designed to communicate, verify and execute contracts in an information-based manner. Smart contracts allow for trusted transactions without a third party, which are traceable and irreversible.

[0026] Consensus Mechanism: Also known as consensus algorithm, it is a process of verifying and confirming transactions in a short period of time through voting by special nodes. For a transaction, if several nodes with irrelevant interests can reach a consensus, it can be considered that the entire network can also reach a consensus on it.

[0027] Zero-Knowledge Proof (ZKP) technology: A cryptographic technique used to prove that a claim is correct without revealing the content of the claim. Specifically, zero-knowledge proves that a person has certain information or that a transaction is valid without revealing the specific information and transaction details. This technology can verify the authenticity of information or transactions while protecting privacy.

[0028] Homomorphic Encryption (HE) technology: refers to an encryption algorithm that satisfies the homomorphic operation property of ciphertext. That is, after the data is homomorphically encrypted, a specific calculation is performed on the data. The ciphertext calculation result obtained after the corresponding homomorphic decryption is equivalent to directly performing the same calculation on the inscription data, thus achieving "calculated but invisible" data.

[0029] Secure Multi-Party Computation (MPC) technology: It is a privacy computing implemented in pure software based on a variety of cryptographic technologies. Secure multi-party computing means that in the absence of a trusted third party, multiple participants collaborate to calculate an agreed function, and ensure that each party only obtains its own calculation results, and cannot infer the input and output data of any other party through the interactive data during the calculation process (unless the function itself can infer the input and output of other participants from its own output).

[0030] Differential Privacy: It is a method in cryptography that aims to maximize the accuracy of data queries when querying from a statistical database while minimizing the chance of identifying its records.

[0031] Proxy Re-encryption (PRE) technology: It is an encryption technology that allows a third party (agent) to modify the ciphertext encrypted for a certain party (data provider) so that other parties (data users) can decrypt the ciphertext. Proxy re-encryption technology includes the following two features: delegation, which allows the message receiver (key holder) to generate a re-encryption key based on its own key and the key of the delegated user. The proxy uses this re-encryption key as input and converts the ciphertext into the key of the delegated user through the re-encryption function; transitivity, that is, the transitive proxy re-encryption scheme allows the ciphertext to be re-encrypted an unlimited number of times. Therefore, proxy re-encryption technology is often used for secure data sharing in cloud environments.

[0032] PoW (Proof of Work): Nodes within the blockchain compete for the right to write blocks on the blockchain by solving complex mathematical problems (usually hash operations). Since these problems require a lot of computing power, only nodes with sufficient computing power can successfully solve them and obtain the right to write blocks.

[0033] PoS (Proof of Stake): The weight of a node in a blockchain is determined by the amount and time of the coins it holds. The more coins a node holds and the longer it holds them, the greater its weight on the blockchain and the higher the probability of obtaining block write rights.

[0034] Example 1

[0035] According to an embodiment of the present application, a data security management method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0036] Figure 1 is a flow chart of a data security management method provided according to an embodiment of the present application, such as Figure 1 As shown, the method comprises the following steps:

[0037] Step S102: obtaining a first data set related to a target field and a target object within a target time period, and obtaining a personalized privacy protection requirement of the target object.

[0038] In the technical solution provided in the above step S102, the above target field refers to a specific industry, research field or application field, such as medical, finance, education and other specific fields. The above first data set includes the first data related to the target field that the target object expects to protect at multiple times within the target time period and the description information of each first data, wherein the description information includes but is not limited to: the data source of the first data, the expected application scenario, etc. For example, the medical data management system can obtain all data related to medical health in the past year from the electronic health record (EHR) system of patient A, including but not limited to: medical records, examination results, medication information, living habits, etc. This data set constitutes the "first data set". In the acquisition process, the system not only collects the first data (i.e., specific data items, such as blood pressure values, blood sugar levels, etc.), but also obtains the description information of each first data, including the source of the data, the collection time, the type (such as whether it is sensitive data), etc., for subsequent data sensitivity analysis.

[0039] In addition, the personalized privacy protection needs of the target object can be understood as the privacy protection preferences of the target object for the first data set. For example, patient A may require the medical data management system to provide the highest level of protection for the disease diagnosis information in his medical records, while setting a lower level of protection for general living habits data. These personalized privacy protection needs can be customized directly by the target object on the interactive interface and recorded by the medical data management system as an important basis for the subsequent selection of privacy protection technology.

[0040] Step S104: Analyze each first data and its description information in the first data set by using the target data sensitivity analysis model to determine the sensitivity level of each first data in the first data set.

[0041] In the technical solution provided in the above step S104, the above target data sensitive analysis model is obtained by fine-tuning the pre-trained initial data sensitive analysis model using the corpus of the target domain. The initial data sensitive analysis model is obtained by training the neural network model with a general corpus (i.e., a large number of general data sets), wherein these general data sets are usually large in scale and cover a wide range of topics and content, such as Wikipedia, Common Crawl, etc. By training on these data, the model can learn the universal laws and rich feature representations of the language, and establish a basic understanding of the language structure and vocabulary usage. Therefore, on the basis of the above pre-trained initial data sensitive analysis model, the corpus (smaller data set) of the target domain is used for further training to adapt to and optimize the performance of the model in this field. For example, a model that has been pre-trained on a general corpus can improve its understanding of medical terms and concepts by fine-tuning on medical text data.

[0042] Based on the above-mentioned target data sensitivity analysis model, the management system can extract features of each first data and its description information in the first data set. These features include but are not limited to: specific vocabulary, numerical range, sensitive information in the image, etc.; then, the target data sensitivity analysis model can further analyze the extracted features to identify the sensitive information that may be contained therein, and determine the corresponding sensitivity level based on the characteristics of the sensitive information. Generally, the sensitivity level can be divided according to the needs of the actual application scenario, such as high, medium and low levels.

[0043] Compared to the existing models that only consider the data itself when classifying data sensitivity levels, the target data sensitivity analysis model provided in the embodiment of the present application is based on considerations of multiple dimensions, including but not limited to the type of data (personal information, financial information, health status, etc.), the source of the data (direct user input, third-party services, sensor data, etc.) and the potential sensitivity of the data (such as whether it contains personal identification information, whether it is easy to be abused, etc.), thereby providing a more comprehensive and accurate data sensitivity assessment. For example, even if it is a seemingly ordinary text data, if it comes from a bank account, involves financial information, and is frequently accessed, then its sensitivity level may be judged to be high. This comprehensive assessment can avoid misjudgments that may occur when evaluating only the data content, ensuring that important privacy information is fully protected.

[0044] Step S106 , determining a target privacy protection strategy for each first data according to the personalized privacy protection requirement of the target object and the sensitivity level of each first data in the first data set.

[0045] In the technical solution provided in the above step S106, the management system combines the personalized privacy protection needs of the target object and the sensitivity level of each first data in the first data set to comprehensively consider multiple factors, including but not limited to the sensitivity level of the data, user privacy preferences, data usage, compliance requirements and technical feasibility, etc., to formulate a personalized and technically feasible target privacy protection strategy, which can not only effectively protect data security, but also meet the personalized needs of users, improve resource utilization efficiency, and ensure that the operation of the system complies with regulatory requirements. Intelligently decide on the target privacy protection strategy for each first data.

[0046] Step S108: encrypt each first data according to the target privacy protection strategy, and store the encrypted first data on the blockchain network.

[0047] In the technical solution provided in the above step S108, after determining the target privacy protection strategy of each first data through the above steps, each first data is encrypted according to the target privacy protection strategy to ensure that sensitive data is protected at the highest level. In addition, since blockchain technology can provide a decentralized, distributed and tamper-proof database, when the encrypted data is stored on the blockchain network, each data block will be timestamped and linked to the previous block to form a continuous chain. This structure ensures that the data cannot be modified or deleted once stored, improves the integrity and persistence of the data, and provides additional protection for data security. At the same time, since another important feature of blockchain technology is transparency and traceability, the data stored on the blockchain is visible to all network participants (although the content is encrypted), but anyone can verify the integrity of the data and the legitimacy of the transaction without revealing the specific content of the data. This increases the transparency of the system, and also facilitates audits and compliance checks to ensure the fairness and compliance of the data processing process.

[0048] Based on the scheme defined in the above steps S102 to S108, it can be known that in the embodiment, the management system uses integrated learning combined with deep learning models and machine learning algorithms to conduct in-depth analysis and sensitivity identification of data, thereby achieving flexibility and customization of privacy protection measures. At the same time, users can dynamically adjust privacy protection measures according to their personal privacy preferences. This flexibility allows users to choose appropriate privacy protection technologies when processing different types of transactions, thereby effectively meeting individual needs. This highly customized solution enhances users' control over privacy protection and improves user satisfaction.

[0049] The following describes each step of the data security management method in conjunction with the specific implementation process.

[0050] As an optional implementation, in the technical solution provided in the above step S102, the management system can obtain the personalized privacy protection needs of the target object through the following method, including: responding to multiple privacy protection parameter setting instructions initiated by the target object on the privacy protection setting interface, and using the privacy protection parameters corresponding to the multiple privacy protection parameter setting instructions as the personalized privacy protection needs of the target object.

[0051] In the above embodiment, a highly interactive web-based user interface (i.e., "privacy protection setting interface") is developed in advance, which can use a front-end framework (such as React or Vue.js, etc.) to create dynamic forms, sliding selectors and other components to provide an intuitive operation experience. At the same time, an advanced graphics library (such as D3.js) can be added to the highly interactive user interface to display the effects of different privacy settings in real time, such as the potential risk of data exposure and the expected privacy protection level.

[0052] The target object can then select the data protection level and technology (i.e. “privacy protection parameters”) that he or she desires through sliders, drop-down menus, checkboxes and other operating instructions in a pre-developed highly interactive user interface. These settings include but are not limited to:

[0053] (1) Privacy protection level: Users want to set different privacy protection levels, such as "high", "medium", and "low", which reflects the degree of control users have over how their data is accessed, used, or shared. For example, for financial data, users may choose a "high" level of protection, while for public comments, they may choose a "low" level of protection.

[0054] (2) Data access control: Users may wish to specify who can access their data and under what conditions. For example, users may only allow specific medical personnel to access their medical records during treatment, or allow data to be used for research only with explicit consent.

[0055] (3) Data retention period: Users may have specific requirements for how long data should be retained in the system before it should be deleted or anonymized. For example, a user may choose to retain their shopping history for one year to facilitate personalized recommendations, after which the data should be anonymized.

[0056] (4) Data sharing scope: Users can set the scope of data sharing, such as sharing personal photos only with friends on social media without making them public to everyone. In professional applications, users may limit data sharing to only within an internal team or among specific business partners.

[0057] (5) Technology preference: Users may have preferences for specific privacy protection technologies, such as wanting to use zero-knowledge proof to verify the legitimacy of their data without directly revealing the data content, or preferring to use homomorphic encryption to ensure the privacy of data during processing.

[0058] (6) Transparency and reporting: Users may want to understand how their data is used and require the system to provide reports on data usage and access, as well as notifications of any potential privacy risks or breaches.

[0059] In addition to the six items listed above, personalized privacy protection requirements may also include other requirements such as transparency and compliance. Since they are not the focus of protection of this application, they will not be listed here one by one. The above-mentioned personalized privacy protection requirements can usually be stored in a personalized privacy configuration file, so that the system can dynamically adjust the privacy protection strategy according to the user's privacy needs when processing other data of the user again in the future. Among them, regarding the personalized privacy configuration file, it can be quickly set up according to the preset personalized privacy protection parameter template, and the multiple requirements configured through the user interface can be filled in the corresponding positions in the template to obtain a personalized privacy requirements file that reflects user needs. The system can automatically update the corresponding personalized privacy protection parameter template by regularly analyzing the feedback information of each user to adapt to the new personalized privacy protection needs.

[0060] In order to further enhance the user experience, ensure that the user's personalized privacy preferences can be correctly implemented and enhance the user's trust in the system, the management system can also integrate an intelligent help system, which can help users understand different privacy settings by providing easy-to-understand guidance to ensure that users can choose the correct settings, so that users can easily set their privacy preferences and reduce setting errors; at the same time, considering that different users may have very different needs and concerns about privacy, the intelligent help system can provide personalized suggestions based on the user's specific needs and scenarios. Specifically, the intelligent help system can use deep learning algorithms to analyze the user's historical setting data (including historical setting parameters, common configuration modes, etc.) on a highly interactive user interface to automatically provide personalized setting suggestions and operation guidance.

[0061] As an optional implementation, due to the data types of the first data in the first data set, these different types of data have different formats and structures. In order to improve the accuracy of the model analysis result, the management system can pre-operate each first data in the first data set according to the following method, including:

[0062] First, determining the data type of each first data in the first data set, wherein the data type includes at least one of the following: text type, voice type, image type;

[0063] In a case where the type of the first data is text type data, the first data is preprocessed using a first preprocessing operation.

[0064] Among them, the above-mentioned first preprocessing operation includes but is not limited to: text cleaning (i.e., removing irrelevant characters such as punctuation marks, numbers, and special symbols; at the same time, regular expressions or natural language processing NLP libraries can be used to filter out non-text content such as HTML tags, URLs, Emails, etc.), case conversion (i.e., converting all text to a unified format, usually choosing lowercase to reduce the difficulty of model learning), word segmentation (i.e., splitting text into words or phrases), removing stop words (i.e., removing common words that are not very meaningful for analysis from the text, such as "de", "he", "zai", etc.), stemming and lemmatization (i.e., reducing words to their basic forms), word encoding (i.e., converting the vocabulary in the text into digital representations, such as Bag-of-Words, TF-IDF, word embeddings), etc.

[0065] When the type of the first data is of the voice type, the first data is preprocessed using the second preprocessing operation.

[0066] Among them, the above-mentioned second preprocessing operation includes but is not limited to: sampling rate normalization (i.e., making the sampling rates of all audio files consistent), framing (i.e., splitting continuous audio signals into short-time windows (frames), usually 20 - 30 milliseconds), noise cancellation (i.e., reducing the influence of background noise through noise suppression algorithms), pre-emphasis (i.e., reducing the attenuation of high-frequency parts in speech signals through pre-emphasis techniques and enhancing signal quality), speech recognition (i.e., converting speech data into text data), etc.

[0067] When the type of the first data is of the image type, the first data is preprocessed using the third preprocessing operation.

[0068] Among them, the above-mentioned third preprocessing operation includes but is not limited to: size adjustment (i.e., adjusting all images to a unified size, such as adjusting all images to 224x224 pixels), data augmentation (i.e., increasing the diversity of image data through operations such as rotation, flipping, scaling, etc.), normalization processing (i.e., normalizing the pixel values of images, such as converting RGB values from 0 - 255 to the range between 0 - 1), and normalization (i.e., normalizing the images to reduce the differences between pixel values).

[0069] By selecting different preprocessing methods according to the data type, we can ensure that the data adapts to the model input format and improve the efficiency and accuracy of model training. In addition, the above preprocessing process can be highly automated by writing customized scripts or using ready-made preprocessing libraries (such as NLTK, Scikit-learn, and TensorFlow's preprocessing modules) to reduce human errors and increase processing speed. It should be noted that the preprocessing operation can be fine-tuned according to the specific model and application scenario to ensure that the preprocessed data can best serve subsequent analysis and learning tasks.

[0070] In addition, since different types of data have significant differences in structure, characteristics, and processing requirements, the architecture of models for analyzing different types of data is also different.

[0071] Specifically, text data is usually presented as a one-dimensional sequence, and each element can be a character, word, or word embedding. Therefore, the processing of text data usually requires a model that can analyze the sequence dependencies and contextual information in the text. Speech data is also a sequence of data, but unlike text data, it includes time series audio signals. Therefore, the processing of speech data usually requires a model that can capture time series features.

[0072] Therefore, when the type of the first data is text type data or speech type, the model architecture of the initial data-sensitive analysis model includes at least one of the following: Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), Recurrent Neural Network (RNN), Transformer model based on attention mechanism (such as BERT), Gated Recurrent Unit (GRU).

[0073] In addition, image data exists in the form of a two-dimensional pixel matrix, and the features in the image often have local correlation and spatial structure. Therefore, the processing of image data usually requires a model that can effectively identify and extract the local features and spatial hierarchical structure of the image.

[0074] When the type of the first data is image type data, the model architecture of the initial data sensitive analysis model includes: a deep residual network (ResNet), a densely connected convolutional network (DenseNet) or a very deep neural network without residual (FractalNet).

[0075] After determining the model architecture of the initial sensitive data analysis model, the pre-trained initial data sensitivity analysis model can be fine-tuned using the corpus of the target domain to obtain a target data sensitivity analysis model that can accurately analyze the sensitivity of data in the target domain. The corpus of the target domain includes: historical data related to the target domain at historical moments of other users and their description information, sensitivity level labels corresponding to each historical data, and the description information includes at least one of the following: data source, data usage scenario.

[0076] In addition, the management system can also regularly obtain sensitive data from authoritative data sources to establish a compliance information sensitivity database that meets industry standards, and use this information sensitivity database to continuously update the model's sensitivity rating standards to ensure the accuracy of the model analysis results.

[0077] After completing the target data sensitivity analysis model, the management system can use containerization technology such as Docker to deploy the model to ensure rapid iteration and adapt to new data formats and privacy requirements. Based on the deployed target data sensitivity analysis model, the management system can call the target data sensitivity analysis model to analyze each first data in the first data set and its description information to determine the sensitivity level of each first data in the first data set.

[0078] For example, the medical data management system obtains all data related to medical health in the past year from the electronic health records (EHR) system of patient A (i.e., the "first data set"), including but not limited to: medical records, examination results, medication information, living habits, etc. Then, these electronic health record data are preprocessed, including removing irrelevant information, standardizing data formats, and performing desensitization processing (such as replacing or deleting personal identification information). Then, the key features of each electronic health record data at different periods of the year are extracted, such as keywords (such as disease names, drug names), digital features (such as age, blood pressure values), date information (such as date of medical treatment), etc., as well as descriptive information of the data, such as source institutions, data types (structured or unstructured), etc. Finally, the target data sensitivity analysis model is used to evaluate the sensitivity level of each data according to the data features and data descriptive information. For example, the medical record text containing personal medical history and diagnosis results is highly sensitive data, while the post-illness recovery opinions on healthy diet recommendations may be marked as low-sensitivity data.

[0079] As an optional implementation, in the technical solution provided in the above step S106, after obtaining the sensitivity level of each first data in the first data set and the personalized privacy protection requirements of the target object through the above steps, the management system can decide a target privacy protection strategy that not only meets user needs but also matches the sensitivity level of each first data itself according to the following method, including: using a pre-trained target decision model to analyze the sensitivity level of the first data and the personalized privacy protection requirements of the target object to obtain the target privacy protection strategy of the first data. Among them, the above target privacy protection strategy includes at least one privacy protection technology, and the privacy protection technology includes but is not limited to: zero-knowledge proof technology, homomorphic encryption technology, secure multi-party computing technology, differential privacy technology, proxy re-encryption technology, etc.

[0080] Specifically, the training process of the above target decision model includes:

[0081] Step 1: Obtain a training sample set and a sample label set. The training sample set includes: multiple training samples consisting of at least sensitivity level labels (such as high, medium, low, etc.) corresponding to historical data related to the target field at historical moments of other users, and personalized privacy protection requirements of other objects that uploaded historical data; the sample label set includes the optimal privacy protection strategy label corresponding to each training sample, wherein the optimal privacy protection strategy label refers to the privacy protection strategy with the highest satisfaction level in the corresponding user feedback after the training samples are encrypted and protected through expert experience and historical data analysis.

[0082] Step 2: Construct an initial decision model, wherein the initial decision model includes at least one of the following: a decision tree model, a logistic regression model, and a support vector machine model.

[0083] Step 3: Use the training sample set and sample label set to iteratively train the initial decision model to obtain the target decision model.

[0084] Specifically, when the above-mentioned initial decision model is a decision tree model, since the decision tree model is a classification model of a tree structure, it makes decisions by constructing a decision tree. Therefore, in this scenario of the embodiment of the present application, the decision tree model can realize decision-making in a process: First, the model automatically selects the most discriminating feature from a series of possible features (such as data sensitivity level, data type, user preference, data usage scenario, etc.) to make decisions. For example, it may first make decisions based on the data sensitivity level (high, medium, low), and then further classify according to the data type (text, image, audio, etc.). Then, at each decision node, the model will determine which child node the data should enter based on the feature value selected on the node. For example, if the data sensitivity level is "high", the model may further check whether the user has selected "anonymous processing" as a privacy preference, and then decide whether to use homomorphic encryption or differential privacy technology. Finally, the data starts from the root node and moves down along the path that meets the conditions until it reaches the leaf node, which represents the final privacy protection strategy. For example, a decision path may be: "Data sensitivity level is high> user preference is anonymous processing> use zero-knowledge proof".

[0085] When the above-mentioned initial decision model is a logistic regression model, since logistic regression is a linear model, it is mainly used to estimate the probability of an event. Therefore, in this scenario of the embodiment of the present application, the logistic regression model can implement decision-making in a process: First, the logistic regression model needs to select and process features, but it focuses more on quantitative features, such as converting data sensitivity levels into numerical levels and converting user privacy preferences into binary or ordinal variables. Then, the logistic regression model will calculate the probability of applying a certain privacy protection strategy (such as "homomorphic encryption") based on the input features. For example, if the data sensitivity level is high and the user prefers strong privacy, the model may give a high probability of applying "homomorphic encryption". Finally, by setting a threshold, the system can convert the probability into a privacy protection strategy for actual application. For example, if the probability of applying "homomorphic encryption" is higher than 80%, the system will apply the strategy to the data.

[0086] When the above-mentioned initial decision model is a support vector machine model, support vector machine (SVM) is a machine learning model for classification and regression, which is particularly suitable for processing data with high-dimensional feature space. Therefore, in this scenario of the embodiment of the present application, the support vector machine model can process decision-making: First, the SVM model processes multidimensional features, including data sensitivity level, user privacy preference, data type, industry, regional regulations, etc. Then, SVM finds an optimal decision boundary by maximizing the boundary distance between different categories to classify data into different privacy protection strategies. For example, it may classify the combination of "highly sensitive data + strong privacy preference" into the "use homomorphic encryption" strategy, while the combination of "lowly sensitive data + weak privacy preference" may be classified into the "use MD5 hash" strategy.

[0087] In addition, the management system can adjust model parameters through the "rewards" obtained from different decisions (such as user satisfaction, system performance and other indicators), thereby continuously improving model performance.

[0088] As an optional implementation, in the technical solution provided in the above step S108, the management system may perform security protection on each first data in the first data set according to the following method, including:

[0089] Step S1081, encrypting the first data according to the target privacy protection policy, and packaging the encrypted first data into a transaction data, wherein the transaction data at least includes: the first data, a transaction timestamp, and a public key of the target object.

[0090] Specifically, before encryption, the management system can also standardize the format of the first data to ensure that the data meets the data format requirements of the blockchain platform, such as converting all data into JSON format. At the same time, in order to increase the data transmission speed, the data can also be compressed according to the data size to reduce storage costs and speed up upload speed. Then, the management system can encrypt the processed first data according to the target privacy protection strategy to ensure the security of the data during transmission and storage; finally, create a transaction containing data or data pointers (such as hash values). It should be noted that the above transaction data can also include smart contract code to automatically execute specific privacy protection rules or data usage logic in the future.

[0091] Step S1082: Use the public key of the target object to sign the transaction data, and broadcast the signed transaction data to each node in the blockchain network.

[0092] Specifically, the transaction data is signed using the private key (also known as the public key) of the sender (i.e. the target object) to prove its source and ensure the integrity of the data, which is the basis of the security of blockchain transactions; the signed transaction is then broadcast to the blockchain network, which is usually done through the blockchain's P2P network to ensure that all nodes can receive the transaction information.

[0093] Step S1083, for each node on the blockchain network, the signed transaction data is verified using a consensus mechanism, and the verified transaction data is added to the block to be packaged, where the block to be packaged also includes other transaction data and a hash value for linking the previous block; the blocks to be packaged are packaged into a new block, and after the new block is confirmed by other nodes on the blockchain network through a consensus mechanism, the new block is added to the blockchain network to update the ledgers of all nodes in the blockchain network.

[0094] In the technical solution provided in the above step S1083, if any node on the blockchain network receives the transaction, it can verify the signed transaction data according to the consensus mechanism of the blockchain (such as PoW, PoS, etc.), and the verification content usually includes: checking the signature of the transaction, ensuring that the format of the data is correct, and executing any necessary smart contract logic. Then, the verified transaction is added to the block to be packaged, which may also contain other transactions and a hash value for linking the previous block. Then, a node (called a verifier) ​​packages the verified transaction into a new block, and then obtains confirmation from other nodes in the network through a consensus mechanism. Once the new block is confirmed by the network, it will be added to the blockchain, and the ledgers of all nodes on the blockchain will be updated so that each node has a complete copy of the ledger, thereby ensuring the transparency and secure storage of the data.

[0095] In the above process, the packaging and uploading of the first data to the blockchain network requires a high degree of security, efficiency and compliance. The decentralized nature, consensus mechanism and tamper-proof ledger records of blockchain technology provide a strong security guarantee for data. At the same time, through smart contracts and advanced encryption technology, the management system can ensure the legal use and compliance of data while protecting user privacy. It should be noted that the specific implementation of uploading data to the blockchain network may vary depending on the type of blockchain used (such as public chain, consortium chain, private chain) and blockchain platform.

[0096] In addition, the management system of the embodiment of the present application integrates an anomaly detection system to analyze the anomaly score of each user's access behavior, and when the anomaly score exceeds a certain threshold, automatically determines that the corresponding user access behavior is abnormal behavior, thereby automatically warning of potential privacy leakage or data abuse risks.

[0097] Through the above solution, it is not difficult to see that the data security management solution provided by the embodiment of the present application has the following technical advantages:

[0098] (1) Using a deep learning algorithm, comprehensively determining the sensitivity level of each first data in the first data set from multiple dimensions, thereby greatly improving the accuracy of data sensitivity determination;

[0099] (2) The target privacy protection strategy of each first data is comprehensively determined based on the personalized privacy protection needs of the target object and the sensitivity level of each first data in the first data set, which greatly enhances the user's control over data privacy protection. It not only enhances the user's sense of parameters and trust in the management system, but also allows the privacy protection strategy to be set in a targeted manner according to the sensitivity level of the data, thereby protecting user data and reducing unnecessary resource consumption;

[0100] (3) Combine blockchain technology to manage data and realize decentralized identity management of users, reducing the risk of data leakage. At the same time, users can better control their identity information and enhance data security and compliance.

[0101] (4) Provide a variety of privacy protection measures for different data sensitivity levels, so that the protection strategy can be adaptively adjusted according to different application scenarios and needs, greatly enhancing the security and flexibility of the system to cope with ever-changing security threats.

[0102] Example 2

[0103] According to an embodiment of the present application, a data security management device for implementing the data security management method in embodiment 1 is also provided. Figure 2 As shown, the data security management device at least includes: an acquisition module 22, an analysis module 24, a decision module 26 and a security protection module 28, wherein:

[0104] The acquisition module 22 is used to obtain a first data set related to the target field of the target object within the target time period, and obtain the personalized privacy protection requirements of the target object, wherein the first data set includes the first data of the target object at multiple moments within the target time period and the description information of each first data.

[0105] The analysis module 24 is used to use a pre-trained data sensitivity analysis model to analyze each first data and its description information in the first data set to determine the sensitivity level of each first data in the first data set, wherein the data sensitivity analysis model is obtained by fine-tuning the neural network model using a corpus in a specific field.

[0106] The decision module 26 is used to decide the target privacy protection strategy of each first data according to the personalized privacy protection requirements of the target object and the sensitivity level of each first data in the first data set.

[0107] The security protection module 28 is used to encrypt each first data according to a target privacy protection policy and store the encrypted first data on the blockchain network.

[0108] It should be noted that each module in the data security management device in the embodiment of the present application corresponds one-to-one to each implementation step of the data security management method in Example 1. Since a detailed description has been given in Example 1, some details not reflected in this embodiment can be referred to Example 1 and will not be repeated here.

[0109] Example 3

[0110] According to an embodiment of the present application, a computer program product is also provided, which includes a computer program, wherein the computer program implements the data security management method in Example 1 when executed by a processor.

[0111] According to an embodiment of the present application, a non-volatile storage medium is also provided, which includes a stored computer program, wherein the device where the non-volatile storage medium is located executes the data security management method in Example 1 by running the computer program.

[0112] According to an embodiment of the present application, a processor is also provided, which is used to run a computer program, wherein the data security management method in Example 1 is executed when the computer program is running.

[0113] According to an embodiment of the present application, an electronic device is also provided, which includes: a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to execute the data security management method in Example 1 through the computer program.

[0114] Specifically, when the computer program is running, the following steps are executed: obtaining a first data set related to a target field within a target time period of a target object, and obtaining personalized privacy protection requirements of the target object, wherein the first data set includes first data of the target object at multiple moments within the target time period and description information of each first data; using a target data sensitivity analysis model to analyze each first data and its description information in the first data set, and determining the sensitivity level of each first data in the first data set, wherein the data sensitivity analysis model is obtained by fine-tuning a pre-trained initial data sensitivity analysis model using a corpus of the target field, and the initial data sensitivity analysis model is obtained by training a neural network model with a general corpus; determining a target privacy protection strategy for each first data based on the personalized privacy protection requirements of the target object and the sensitivity level of each first data in the first data set; for each first data, encrypting the first data according to the target privacy protection strategy, and storing the encrypted first data on the blockchain network.

[0115] As an optional implementation, the electronic device may be in the form of a mobile terminal, a computer terminal or a similar computing device. Figure 3 FIG. 1 shows a hardware structure block diagram of an electronic device for implementing a data security management method. Figure 3 As shown, the electronic device 30 may include one or more (302a, 302b, ..., 302n are used to illustrate) processors 302 (the processor 302 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 304 for storing data, and a transmission device 306 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the BUS bus), a network interface, a power supply and / or a camera. It can be understood by those skilled in the art that Figure 3 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 3 More or fewer components as shown, or with Figure 3 Different configurations are shown.

[0116] It should be noted that the one or more processors 302 and / or other data processing circuits described above may generally be referred to herein as "data processing circuits". The data processing circuits may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuit may be a single independent processing module, or may be incorporated in whole or in part into any of the other components in the electronic device 30. As described in the embodiments of the present application, the data processing circuit acts as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0117] The memory 304 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the data security management method in the embodiment of the present application. The processor 302 executes various functional applications and data processing by running the software programs and modules stored in the memory 304, that is, realizing the vulnerability detection method of the above-mentioned application. The memory 304 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 304 may further include a memory remotely arranged relative to the processor 302, and these remote memories may be connected to the electronic device 30 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0118] The transmission device 306 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the electronic device 30. In one example, the transmission device 306 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 306 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0119] The display may be, for example, a touch screen liquid crystal display (LCD) that enables a user to interact with a user interface of the electronic device 30 .

[0120] The serial numbers of the above embodiments are only for description and do not represent the advantages or disadvantages of the embodiments.

[0121] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0122] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0123] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed over multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0124] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0125] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or all or part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of each embodiment method of the present application. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk, etc. Various media that can store program codes.

[0126] The above are only preferred implementations of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A data security management method, characterized in that: include: Acquire a first data set related to a target field of a target object within a target time period, and acquire personalized privacy protection requirements of the target object, wherein the first data set includes first data of the target object at multiple moments within the target time period and description information of each of the first data; Analyzing each first data and its description information in the first data set using a target data sensitivity analysis model to determine a sensitivity level of each first data in the first data set, wherein the data sensitivity analysis model is obtained by fine-tuning a pre-trained initial data sensitivity analysis model using a corpus of the target domain, and the initial data sensitivity analysis model is obtained by training a neural network model using a general corpus; Determining a target privacy protection strategy for each of the first data according to the personalized privacy protection requirements of the target object and the sensitivity level of each of the first data in the first data set; For each first data, the first data is encrypted according to the target privacy protection policy, and the encrypted first data is stored on the blockchain network.

2. The method according to claim 1, characterized in that Obtaining the personalized privacy protection requirements of the target object, including: In response to multiple privacy protection parameter setting instructions initiated by the target object on the privacy protection setting interface, the privacy protection parameters corresponding to the multiple privacy protection parameter setting instructions are used as the personalized privacy protection requirements of the target object, wherein the personalized privacy protection requirements include at least one of the following: privacy protection level, data access control, data retention period, and data sharing scope.

3. The method according to claim 1, characterized in that Before analyzing each first data in the first data set and its description information using the target data sensitive analysis model, the method further includes: Determine a data type of each first data in the first data set, wherein the data type includes at least one of the following: text type, voice type, image type; In the case where the type of the first data is the text type data, preprocessing the first data using a first preprocessing operation, wherein the first preprocessing operation includes at least one of the following: text cleaning, case conversion, word segmentation, stop word removal, stemming, and word form restoration; In the case where the type of the first data is the voice type, preprocessing the first data using a second preprocessing operation, wherein the second preprocessing operation includes at least one of the following: frame processing, noise elimination, and voice recognition; In a case where the type of the first data is the image type, the first data is preprocessed using a third preprocessing operation, wherein the third preprocessing operation includes at least one of the following: size adjustment, data enhancement, and normalization.

4. The method according to claim 3, characterized in that In the case where the type of the first data is the text type data or the voice type, the model architecture of the initial data sensitive analysis model includes at least one of the following: a convolutional neural network, a long short-term memory network, a recurrent neural network, and a Transformer model based on an attention mechanism; In the case where the type of the first data is the image type data, the model architecture of the initial data-sensitive analysis model includes: a deep residual network or a densely connected convolutional network.

5. The method according to claim 1, characterized in that Determining a target privacy protection strategy for each first data according to the personalized privacy protection requirements of the target object and the sensitivity level of each first data in the first data set includes: For each first data in the first data set, a pre-trained target decision model is used to analyze the sensitivity level of the first data and the personalized privacy protection requirements of the target object to obtain a target privacy protection strategy for the first data, wherein the target privacy protection strategy includes at least one privacy protection technology, and the privacy protection technology includes at least one of the following: zero-knowledge proof technology, homomorphic encryption technology, secure multi-party computing technology, differential privacy technology, and proxy re-encryption technology.

6. The method according to claim 5, characterized in that The training process of the target decision model includes: Acquire a training sample set and a sample label set, wherein the training sample set includes a plurality of training samples consisting of at least sensitivity level labels corresponding to historical data related to the target domain at historical moments of other users and personalized privacy protection requirements of other objects that uploaded the historical data, and the sample label set includes optimal privacy protection strategy labels corresponding to each of the training samples; Constructing an initial decision model, wherein the initial decision model includes at least one of the following: a decision tree model, a logistic regression model, and a support vector machine model; The initial decision model is iteratively trained using the training sample set and the sample label set to obtain the target decision model.

7. The method according to claim 1, characterized in that Encrypting the first data according to the target privacy protection policy, and storing the encrypted first data on the blockchain network, includes: Encrypting the first data according to the target privacy protection policy, and packaging the encrypted first data into transaction data, wherein the transaction data at least includes: the first data, a transaction timestamp, and a public key of a target object; Sign the transaction data using the public key of the target object, and broadcast the signed transaction data to each node in the blockchain network; For each node on the blockchain network, the signed transaction data is verified by using a consensus mechanism, and the verified transaction data is added to the block to be packaged, wherein the block to be packaged also includes other transaction data and a hash value for linking to the previous block; the blocks to be packaged are packaged into a new block, and after the new block is confirmed by other nodes on the blockchain network through a consensus mechanism, the new block is added to the blockchain network to update the ledgers of all nodes in the blockchain network.

8. A data security management device, characterized in that: include: An acquisition module, used to acquire a first data set related to a target field of a target object within a target time period, and to acquire a personalized privacy protection requirement of the target object, wherein the first data set includes first data of the target object at multiple moments within the target time period and description information of each of the first data; an analysis module, configured to analyze each first data and its description information in the first data set by using a target data sensitivity analysis model, and determine a sensitivity level of each first data in the first data set, wherein the data sensitivity analysis model is obtained by fine-tuning a pre-trained initial data sensitivity analysis model by using a corpus of the target domain, and the initial data sensitivity analysis model is obtained by training a neural network model by using a general corpus; A decision module, configured to determine a target privacy protection strategy for each first data according to the personalized privacy protection requirements of the target object and the sensitivity level of each first data in the first data set; The security protection module is used to encrypt each of the first data according to the target privacy protection policy, and store the encrypted first data on the blockchain network.

9. A computer program product, characterized in that include: A computer program, wherein when the computer program is executed by a processor, it implements the data security management method described in any one of claims 1 to 7.

10. An electronic device, characterized in that: include: A memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the data security management method according to any one of claims 1 to 7 through the computer program.

Images